Skip to content

Opening book details…

About this document

Hybrid AHS for Acoustic Howling Suppression by augustisee is a document available to read on EtoBox.

The document presents a hybrid method for acoustic howling suppression (AHS) that combines a Kalman filter with a self-attentive recurrent neural network (SARNN) to enhance performance during both offline training and streaming inference. The proposed Hybrid AHS method aims to leverage the strengths of traditional adaptive filtering and deep learning to effectively suppress howling while minimizing signal distortion. Experimental results demonstrate that this hybrid approach consistently outperforms existin

Author
augustisee
Language
EN